Next Article in Journal
Biotechnological Potential of Newly Isolated Microalga Strain in Cu and Cr Biosorption from Single and Bimetallic Systems
Previous Article in Journal
A Sediment Process Simulation on the Steep Area of the Upper Yangtze River Basin Using a Hybrid Distributed Soil Erosion Model
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Water
Volume 17
Issue 7
10.3390/w17070998
water-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Mechanisms and Implications of Phosphate Retention in Soils: Insights from Batch Adsorption Experiments and Geochemical Modeling

by
Zhi Tang
1,2,
Zeyong Chi
3,*,
Fengcheng Jiang
4,
Mingzhe Zhao
1,2,
Shengbo Fu
1,2,
Lingqiao Wei
1,2,
Qingsheng Feng
1,2,
Yongming Wu
3 and
Nuchao Xu
5,*
1
Six Geological Team of Hubei Geological Bureau, Xiaogan 432000, China
2
Hubei Key Laboratory of Resources and Eco-Environment Geology, Xiaogan 432000, China
3
Research Institute of Poyang Lake, Jiangxi Academy of Sciences, Nanchang 330012, China
4
School of Resources and Environment, Henan Polytechnic University, Jiaozuo 454003, China
5
Chinese Academy of Environmental Planning, Beijing 100041, China
*
Authors to whom correspondence should be addressed.
Water 2025, 17(7), 998; https://doi.org/10.3390/w17070998
Submission received: 3 March 2025 / Revised: 24 March 2025 / Accepted: 26 March 2025 / Published: 28 March 2025
(This article belongs to the Section Soil and Water)
Browse Figures
Versions Notes

Abstract

:
Soil plays a critical role as a natural barrier in mitigating the infiltration of industrial-derived phosphate pollution into groundwater, with its phosphate retention capacity governed mainly by its mineralogical composition. In this study, three soil samples were collected from the Huangmailing phosphate mine area, and the minerals responsible for phosphate retention were identified through batch adsorption experiments, chemical extraction, and spectroscopy analyses. The distribution of phosphate retention within soil samples was further quantified using a geochemical model. The results indicate that the adsorption capacity of soils to phosphate ranges from 0.193 to 0.217 mg/g. Adsorption equilibrium was reached at 750 min, conforming to the intra-particle diffusion kinetic model. Elevated temperatures facilitate phosphate adsorption. Under acidic and neutral conditions, approximately 80–90% of the phosphate is adsorbed onto iron oxides, primarily through inner-sphere surface complexation, thus unaffected by ionic strength. Under alkaline conditions, the retention mechanism was dominated by the release of exchangeable Ca2+ from vermiculite and biotite, as well as the precipitation of hydroxyapatite. Notably, the critical pH at which the retention mechanism shifts decreased with increasing content of layered silicate minerals and the concentration of cations in the solution. Our study underscores the distinct roles of effective minerals in phosphate retention under different pH conditions and highlights the significance of exchangeable Ca2+ in layered silicate minerals under alkaline conditions. Based on these findings, it is recommended that sites with favorable mineralogical characteristics tailored to the pH of phosphate-containing wastewater be prioritized for phosphorus chemical industries. This study also assesses the cost-effectiveness of adding vermiculite to soil in industrial and agricultural applications. The findings can provide a scientific basis for preventing groundwater phosphorus pollution in critical areas.

1. Introduction

Groundwater is a vital resource, providing drinking water for billions of people worldwide and contributing to 40% of agricultural irrigation water and one-third of industrial water use [1]. However, phosphate pollution poses a significant threat to the quality of groundwater resources, with approximately 10% of global groundwater systems exhibiting phosphate concentrations exceeding 0.1 mg/L [2]. Furthermore, phosphate from groundwater can migrate into surface waters, exacerbating eutrophication and disrupting ecological balance [3,4,5,6]. The production process associated with phosphate fertilizer, including phosphate mining, phosphate chemical enterprises, and phosphogypsum storage facilities, is widely recognized as a significant source of environmental phosphate load [7,8,9]. Reported phosphate concentrations in phosphogypsum leachate can reach tens of thousands of mg/L, with associated groundwater phosphate concentrations as high as 47 mg/L [8,10]. This significant discrepancy is primarily related to the phosphate retention by soil [11,12,13,14]. Since soil serves as a critical barrier to the migration of phosphate, understanding the retention behavior and mechanisms of phosphate in soil is essential for developing effective strategies to mitigate environmental risks associated with phosphate-related industries.
Previous studies have demonstrated that various factors, including contact time, temperature, pH, ionic strength, and coexisting ions, influence phosphate adsorption in soil. Phosphate adsorption in soil typically reaches equilibrium within several hours and can be divided into two stages, described by pseudo-second-order kinetics [15,16,17]. The adsorption capacity of natural soil for phosphate exhibits significant variability, with reported values ranging from 0.05 to 1.00 mg/g [15,17,18,19]. Increasing ionic strength has been shown to promote phosphate adsorption, supporting the inner-surface complexation mechanism [20]. While phosphate adsorption in soil generally decreases with increasing pH, certain calcareous soils demonstrate enhanced adsorption at high pH conditions due to the precipitation of hydroxyapatite under alkaline conditions [15,21,22,23,24]. Additionally, phosphate exhibits a preferential adsorption sequence among anions, indicating that its adsorption in soil is minimally affected by coexisting anions [25].
Notably, mineralogical characteristics significantly influence the adsorption behavior of phosphate in soil, particularly regarding adsorption capacity and pH effects [26,27]. Iron oxides, aluminum oxides, carbonate minerals, and layered silicate minerals are currently considered to have potential contributions to the adsorption of phosphate in soil [6,28,29,30]. Among these, iron oxides, aluminum oxides, and layered silicate minerals exhibit the strongest affinity for phosphate under low pH conditions and are typically considered the primary target phases for phosphate adsorption in acidic environments [15,31,32]. In contrast, carbonate minerals immobilize phosphate through surface precipitation under alkaline conditions [4]. Furthermore, phosphate retention is related to the soluble calcium content in the soil under high pH conditions, primarily due to the precipitation of hydroxyapatite [21]. However, these conclusions are based mainly on chemical extraction and X-ray Absorption Spectroscopy methods, which only provide information on the phosphate distribution in specific samples [6,15].
Geochemical models are effective tools for quantifying phosphate adsorption, as they can simulate adsorption behavior under varying environmental conditions [33]. However, due to the complex mineralogical composition of soils, previous studies have predominantly treated soil as a single entity when developing geochemical models, a method known as the Generalized Composite Approach (GCA) [28,34]. While the GCA can accurately describe the macroscopic adsorption behavior of phosphate in soil, it does not provide information on the distribution of phosphate among different mineral components. In contrast, the Generalized Additivity Approach (GAA) considers soil as a composite of identifiable and quantifiable individual minerals and constructs geochemical models based on the soil’s mineralogical characteristics [35]. This approach enables the observation of phosphate adsorption across different mineral phases but requires prior calibration of surface complexation reaction parameters for each individual mineral. Given that a comprehensive database of surface complexation reactions for phosphate on most potential minerals has been established, the GAA method is a feasible approach for describing the distribution of phosphate in soil [30,31,32].
The key scientific challenge lies in elucidating how phosphate is retained in the soil and determining the proportion of phosphate retained in each form. In this study, we aim to quantitatively assess the retention and distribution of phosphate in soils using the GAA method based on the characterization of the mineralogical properties of the soil samples and observations of phosphate adsorption behavior. This study is expected to advance the understanding and prediction of phosphate migration and fate within soil systems while potentially providing novel insights for environmental risk management strategies in phosphate-related industries.

2. Materials and Methods

2.1. Sample Collection and Preprocessing

The Huangmailing phosphate mine is located in Hubei province, China, approximately 80 km in a straight line from the megacity of Wuhan. This mining area integrates phosphate mining and chemical production, with an annual output of around 600,000 tons of various phosphate fertilizers. Within the mining area, the phosphate chemical industry, the phosphogypsum repository, and the recycling water pond are identified as potential sources of phosphate pollution. The currently operational phosphogypsum repository and phosphate chemical industry in the north region are in the same small watershed. In contrast, the sealed phosphogypsum repository in the south area lies in another small watershed. Surface water from these two watersheds flows into the Huan River, eventually merging into the Yangtze River via the Fuhuan River (Figure 1).
Soil samples were collected in April 2024 from the Huangmailing phosphate mine area. Considering the migration of phosphate pollution towards low-lying areas, the sampling sites were selected downstream from potential contamination sources. Three soil samples were collected, with Sample 1 (S1) collected 50 m downstream from the key risk area—the phosphate chemical industry. Sample 2 (S2) and Sample 3 (S3) were collected from the outlets of two different small watersheds, representing the convergence points of soil migration within the watersheds. As this study does not focus on investigating the soil pollution status of the research area, the non-uniform sampling approach employed was deemed appropriate. The soil samples were collected after removing the top 2 cm of the soil layer, with a sampling depth of 20 cm, and each sample weighed 500 g. After collection, the samples were air-dried in the dark at room temperature and sieved through a 2 mm mesh to remove gravel and other coarse materials.

2.2. Analytical Properties Analysis

Unground samples were used to analyze soil pH, CEC, pHpzc, and particle size distribution. For pH measurement, the soil was mixed with CO2-free deionized water at a solid-to-liquid ratio of 10:25 (w/v). The mixture was vigorously shaken for 2 min and then allowed to stand for 30 min before the soil pH was measured. The CEC was determined using the hexaamminecobalt(III) chloride method, as recommended by the Ministry of Ecology and Environment of China [36]. The pHpzc was measured using an acid-base potentiometric titration method, where soil suspensions (with a solid-to-liquid ratio of 0.56 g:70 mL) were titrated from pH 2.9 to 10.0 with 0.1 M NaOH in the presence of 0.1, 0.3, and 0.7 M NaCl electrolyte backgrounds to identify the pH at which the surface charge is zero. Soil dispersion in water was achieved through ultrasonic treatment, and the particle size of the soil samples was determined using a laser particle size analyzer (Malvern Mastersizer 3000+ Ultra, Malvern Panalytical, Malvern, UK).
The soil samples were partially ground using an agate mortar and sieved through a 200-mesh sieve. Subsequently, the mineralogical and chemical compositions of the soil samples were analyzed using an X-ray diffractometer (Bruker D8 Advance, Billerica, MA, USA, Figure S1), an X-ray fluorescence spectrometer (Rigaku ZSX Primus III+, Showima, Tokyo, Japan, Table S1), and a total organic carbon analyzer (Shimadu TOC-L CPH, Nakagyo-ku, Kyoto, Japan, Table S1). Additionally, the grounded samples were heated at 1000 °C for 3 h to determine the loss on ignition [37].
Free iron oxides (FeEx) in the unground soil samples were extracted using a 0.1 M ascorbic acid + 0.2 M oxalic acid solution (pH 3) [37]. Free aluminum oxides (AlEx) were extracted via the dithionite–citrate–bicarbonate (DCB) method [38]. The concentrations of Fe and Al in the extracted solutions were determined using an Inductively Coupled Plasma-Optical Emission Spectrometer (Thermo iCAP 7600, Waltham, MA, USA).

2.3. Adsorption Batch Experiment

Batch experiments were conducted to investigate the adsorption behavior of phosphate on the three soil samples, examining the effects of contact time, initial phosphate concentration, temperature, solid-to-liquid ratio, pH, and ionic strength. Each soil sample was subjected to eight sets of experiments, with the specific conditions for each set detailed in Table 1. However, Table 1 presents only the average conditions, as manual operations make it challenging to ensure complete consistency across different experiment batches. For instance, the concentration of phosphorus in the desorption experiments was found to be significantly higher than the predetermined 100 mg/L. Simulated solutions with defined background electrolytes and phosphate concentrations were prepared using NaCl, Na2HPO4, and NaH2PO4. The unground soil samples were mixed with the simulated solutions at predetermined solid-to-liquid ratios. The pH of the suspensions was adjusted to the target value using small amounts of NaOH or HCl solution. The mixtures were then agitated at 180 rpm, and samples were collected at predetermined time intervals. At each sampling point, the pH of the suspension was re-measured, and the solution was filtered through a 0.45 µm membrane to obtain a clear filtrate for analysis.
Desorption experiments were conducted based on the isothermal adsorption experiments. In these experiments, the soil was first equilibrated with a 100 mg/L phosphate solution and then diluted with a 0.01 M NaCl solution to initiate desorption. The solid-to-liquid ratio for desorption experiments was adjustable, depending on the dilution fold. Since adsorption equilibrium is governed by the balance of phosphate concentrations between the liquid and solid phases, variations in the solid-to-liquid ratio did not affect the substantive conclusions derived from the desorption experiments.
For each experiment, a set of parallel experiments was conducted simultaneously to minimize random errors. Additionally, a blank group was prepared simultaneously to avoid interference from the leaching of phosphate naturally present in the soil. The blank group was done by conducting batch experiments with electrolyte solutions that did not contain phosphate. Phosphate concentration was determined using a microplate reader (SpectraMax iD3, San Jose, CA, USA) through the ammonium molybdate spectrophotometric method. In the presence of acid and antimony salt, phosphate reacts with ammonium molybdate to form phosphomolybdic heteropolyacid, which is then reduced by ascorbic acid to produce a blue complex. During sample measurement, 10% of the samples were randomly inserted as replicates to ensure data reliability, with the deviation of replicate measurements typically less than 5%. The amount of phosphate adsorbed by the soil was calculated using the following formula:
Q = V × C i − ( C e − C 0 ) M s
where Q is the amount of phosphate adsorbed per unit mass of soil (mg/g), V represents the volume of the phosphate solution (L), Ci represents the initial phosphate concentration (mg/L), Ce represents the equilibrium phosphate concentration (mg/L), C0 represents the phosphate concentration of the equilibrium solution in the blank group (mg/L), and Ms represents the mass of the soil (g).

2.4. Characterization of the Adsorption Mechanism

The adsorption mechanisms of phosphate in soil under different environmental conditions were investigated by comparing the spectroscopic characteristics of soil samples before and after phosphate adsorption. Three distinct environmental conditions were established based on the results of batch adsorption experiments: (1) equilibration with a 0.05 M NaCl solution devoid of phosphate; (2) equilibration under acidic conditions with a solution containing 0.05 M NaCl and 100 mg/L phosphate; (3) equilibration under alkaline conditions with a solution containing 0.05 M NaCl and 100 mg/L phosphate. Other conditions for sample preparation were consistent with those used in the batch adsorption experiments. After equilibration, the samples were filtered, dried, and ground using an agate mortar to pass through a 200-mesh sieve. The prepared samples were mixed with dry chromatographic-grade KBr and pressed into piece for analysis using a Fourier Infrared Spectrometer (Thermo Fisher Scientific Nicolet iS20, Waltham, MA, USA). The samples after phosphate adsorption were also tested using an X-ray Diffractometer (Bruker D8 Advance, Billerica, MA, USA) to further analyze the mineralogical changes associated with phosphate adsorption.

2.5. Geochemical Model

Given that this study aims to serve the prediction of phosphate migration behavior in the environment, we selected the PHREEQC Interactive 3.7.3 code, which is more compatible to establish a geochemical model that describes the retention and distribution of phosphate in soil. This choice allows the model’s later application in solute transport simulation software. Specifically, the GAA method was used to construct the geochemical model, enabling the model to identify the details of the interactions between different minerals and phosphate. The surface complexation model was constructed using a non-electrostatic module, with the primary parameters being the equilibrium constants, specific surface area, and surface site density. Experimental data on phosphate adsorption on effective minerals were obtained from the previous studies [31,32], and the equilibrium constants were refitted using the PHREEQC Interactive 3.7.3 code to ensure consistency and avoid discrepancies due to software compatibility issues (Figure S2). Given the complexity of the forms of free iron and aluminum oxides in soil, the specific surface area and surface site density parameters were selected within the reported range of values to optimize the fitting results of the geochemical model. The geochemical model exhibits high sensitivity to variations in equilibrium constants, moderate sensitivity to specific surface area and site density, and significant changes in output results can still be observed with a 10% variation in specific surface area and site density (Figure S3). The model also accounted for the influence of soluble Ca2+ on phosphate retention in the soil by measuring the content of dissolved cations in the soil (Table S2). The supporting materials detail the methods for determining the content of soil-soluble cations.

3. Results

3.1. Soil Characteristics

As shown in Figure S1 and Table S1, the soil samples mainly contain inorganic minerals such as quartz, albite, orthoclase, vermiculite, and biotite, with a total organic carbon content between 1.26% and 1.78%. Notably, S3 contains more orthoclase than the other two samples but has lower albite content and lacks vermiculite, likely due to its origin from a different small watershed. The pH of S1 and S2 are 7.22 and 7.28, respectively, classifying them as neutral soils, while the pH of S3 is 6.10, which is considered weakly acidic. The pHpzc for S1, S2, and S3 are 8.90, 9.40, and 7.69, respectively, and their CEC are 7.1, 5.8, and 1.0 cmol/kg, respectively. The clay content of S1, S2, and S3 is 2.48%, 6.63%, and 4.25%, respectively, with median particle sizes of 0.0627, 0.0309, and 0.0386 mm, indicating that S2 may have a higher potential for phosphate adsorption (Figure S4).
Overall, S1 and S2 have similar provenance, leading to comparable physical, chemical, and mineralogical characteristics, while differing significantly from S3. Specifically, the higher CEC of S1 and S2 is likely associated with their higher vermiculite content, as vermiculite is rich in exchangeable cations. Additionally, the higher pHpzc values of S1 and S2 can be attributed to their greater contents of FeEx and AlEx, given that the pHpzc of iron and aluminum oxides typically exceeds that of silicate minerals.

3.2. Adsorption Kinetics

The adsorption of phosphate in soil samples is initially very rapid, but it gradually slows down with increasing contact time, ultimately reaching equilibrium around 750 min. As shown in Table 2, among the three validated models, the intra-particle diffusion model fits the experimental data best, indicating that the adsorption of phosphate by soil is a multi-stage process [39]. The first stage was surface adsorption, the second stage was intra-particle diffusion adsorption, and the last stage encompassed equilibrium adsorption. The diffusion rate constants of three stages were ranked K1 > K2 > K3. In the first stage, the phosphate anions rapidly diffused on the external surface of the soil, and a large number of available sites combined with phosphate formed a concentration gradient between the surface of the soil particle. The phosphate adsorption rate was highest during the first process. The adsorption rate began to decrease during the second stage due to the adsorption mainly occurring in the small pores of the soil, and intraparticle diffusion was a major limiting factor. In the last stage, the adsorption is near equilibrium due to the extremely low phosphate concentration and fewer available sites. These curves did not pass through the origin, which suggested that the intraparticle diffusion was not the only rate-controlling step, while the film diffusion and intraparticle diffusion were synchronized in the adsorption course [40].

3.3. Isothermal Adsorption

The Langmuir, Freundlich, and Sips models all effectively fitted the isotherm adsorption curves of phosphates in soil (Figure 2). In most cases, the R2 value for the Freundlich model was higher than that for the Langmuir and Sips model, indicating that the adsorption of phosphates in the soil is more consistent with multilayer adsorption on heterogeneous sites [41]. This phenomenon may be related to the complex variety of active sites in the soil or the strong heterogeneity in the spatial distribution.
The Langmuir fitting results showed that the adsorption capacities for phosphates by S1, S2, and S3 were 0.202, 0.217, and 0.193 mg/g, respectively, which fall within the reported range of adsorption capacities.

3.4. Adsorption Thermodynamics

The adsorption amount of phosphate in all three soil samples increased with the rise in temperature, indicating that phosphate adsorption is more favorable at higher temperatures. As shown in Table 3, the ΔG0 values for phosphate adsorption were all negative, signifying that the adsorption reaction is favorable and spontaneous [42]. The ΔG0 value for S2 was significantly lower than those for S1 and S3, suggesting that the adsorption of phosphate on S1 and S3 is thermodynamically more favorable than on S2, which is consistent with the results of the kinetic experiments.
The ΔG0 values generally exhibited a decreasing trend with increasing temperature, further supporting that the adsorption of phosphate in the soil is thermodynamically favorable at higher temperatures. The positive ΔH0 value indicated that the adsorption of phosphate in soil is inherently an endothermic reaction. The positive ΔS0 value reflected that the adsorption of phosphate in the soil is an entropy-increasing process, which may be attributed to the desorption of water molecules from the soil particle surfaces during phosphate adsorption [43].

3.5. Desorption

In this study, the soil was first pre-equilibrated with a high concentration of phosphate, followed by dilution of the liquid-phase phosphate concentration using an electrolyte solution to achieve desorption. Since factors directly affecting adsorption equilibrium, such as temperature and background electrolyte, remained constant, comparing the desorption curve with the adsorption curve provides insights into the reversibility of the adsorption reaction. The desorption curve of phosphate on the soil was slightly higher than the adsorption curve, indicating that phosphate adsorption is partially reversible (Figure 3). This behavior is consistent with the desorption characteristic of phosphate on iron and aluminum oxides. However, the soil exhibited a higher desorption rate (20–31%) compared to ferrihydrite, goethite, hematite, and aluminum oxide (5–20%) [44,45]. This difference may be attributed to the higher pre-equilibration phosphate concentration used in this study.

3.6. Effect of Soil Dose on Phosphate Adsorption

The concentration of phosphate in the liquid phase decreased with the increase in soil dosage (Figure 4). Under the condition of a 3 g soil dosage, the removal rates of phosphate from the solution by S1, S2, and S3 were 15.7%, 24.5%, and 16.2%, respectively, indicating a better phosphate affinity for S2, which is consistent with the results from the isothermal adsorption experiments (Figure 2). However, the phosphate load on the soil decreased with the increase in soil dosage, primarily because, under the same conditions of pH, temperature, and background electrolyte, the amount of phosphate adsorbed in the solid phase depends only on the concentration of phosphate in the liquid phase at the equilibrium of the reaction, and not on the solid-to-liquid ratio. Regardless, the solid-to-liquid ratio in natural soil environments is much higher than that in the batch experiments conducted in this study, which means that the soil system has a stronger capability to impede the migration of exogenous phosphate into groundwater.

3.7. Effect of pH and Ionic Strength on Phosphate Adsorption

Under conditions ranging from pH 3.0 to 8.5, the adsorption of phosphate in soil decreases with increasing pH, consistent with reported behaviors of phosphate adsorption on minerals such as iron oxides and aluminum oxides (Figure 5) [13,15,45,46]. This trend is primarily because phosphate species with a lower degree of dissociation exhibit a stronger adsorption affinity. The adsorption of phosphate on samples S1 and S2 increases sharply around pH 8.5, aligning with the behavior observed in calcareous soils [11]. In contrast, the adsorption of phosphate on S3 continues to decrease across the entire pH range. However, XRD results showed that no carbonate minerals (such as calcite, dolomite, or aragonite) were present in any of the soil samples, indicating that the enhanced retention of phosphate in S1 and S2 under alkaline conditions is not driven by the same mechanisms as those in calcareous soils.
Below pH 8.5, the influence of ionic strength on the adsorption of phosphate on soils is negligible, which affirms that the predominant adsorption mechanism is inner-sphere surface complexation, as opposed to physical processes like electrostatic interactions [47]. However, when the pH exceeds 8.50, the adsorption of phosphate on S1 and S2 under 0.1 M NaCl electrolyte conditions is significantly higher than under 0.01 M NaCl, suggesting that increased ionic strength promotes the retention in S1 and S2 under alkaline conditions. Furthermore, as ionic strength increases, the critical pH for enhanced phosphate retention shifts gradually toward neutrality. In contrast, the adsorption of phosphate on S3 under alkaline conditions is slightly suppressed with increasing ionic strength, which may be attributed to a reduction in the activity coefficient of phosphate in the solution.

4. Discussion

4.1. Target Mineral Phase for Phosphate Adsorption

The target mineral phases for phosphate adsorption in soil remain debated. Some studies suggest that iron oxides and aluminum oxides are the primary contributors to phosphate adsorption in soil, while others highlight a strong relationship between phosphate adsorption behavior and the content of clay minerals [15,22,23,48]. Identifying the main target minerals for phosphate adsorption is crucial for investigating phosphate retention and distribution, as it helps avoid unnecessary or difficult-to-obtain parameters in the construction of geochemical models.
As shown in Figure 6, the phosphate adsorption capacity of the soil exhibited a good positive correlation with the contents of albite, organic matter, FeEx, and AlEx. Among these, albite is generally not considered an active adsorption component due to its smaller specific surface area and the lack of surface-active sites [49]. Organic matter, on the other hand, serves as a primary source rather than a sink of phosphate in the soil system through the mineralization processes [50]. Biotite and vermiculite do not show a strong correlation with the phosphate adsorption capacity of the soil samples. These findings suggest that FeEx and AlEx are the main target mineral phases for phosphate adsorption in the soil samples under isothermal adsorption experiment conditions (neutral pH).
The ionic strength had no significant effect on phosphate adsorption under acidic conditions, supporting the hypothesis that phosphate adsorption occurs through inner-sphere surface complexation (Figure 5). This finding aligns with previous studies, which suggest that phosphate forms bidentate and monodentate complexes with hydroxyl groups on the surface of iron oxides, while primarily forming monodentate complexes with aluminum oxides [31,32]. When the pH of the liquid phase exceeds the pHpzc, iron and aluminum oxides acquire a negative surface charge, leading to electrostatic repulsion with phosphate. In such cases, an increase in ionic strength theoretically compresses the electrical double layer, reduces electrostatic repulsion, and enhances phosphate adsorption. This behavior is consistent with the observations for S1 and S2 but contrasts with that of S3. Notably, the enhanced phosphate adsorption by S1 and S2 under alkaline conditions deviates from the typical behavior of iron and aluminum oxides reported in previous studies [31,32]. Moreover, although S2 exhibits higher contents of FeEx and AlEx, the trend of enhanced phosphate adsorption under alkaline conditions is less pronounced in S2 compared to S1. These observations suggest that phosphate adsorption in soil is not predominantly governed by electrostatic forces. Under acidic conditions, phosphate adsorption is primarily mediated by the formation of inner-sphere complexes with free iron oxides and free aluminum oxides. In contrast, under alkaline conditions, other minerals or processes play a dominant role in controlling phosphate adsorption.

4.2. Mechanisms of Phosphate Retention in Soil at Alkaline Condition

Since the three soil samples were not calcareous, surface precipitation of phosphate on carbonate minerals could not explain the enhanced retention of phosphate under alkaline conditions. To elucidate the retention mechanism of phosphate in soil under alkaline conditions, a geochemical model was constructed using chemical extraction results to replicate the adsorption behavior of phosphate in three soil samples (Table 4). The model effectively simulates the isotherm curves at 25 °C (Figure 2), with RMSE of 0.0227, 0.0317, and 0.0059 mg/g for S1, S2, and S3, respectively. The RMSE of the geochemical model for the adsorption of phosphate at different soil dosages was 0.0287 mg/g (Figure 4). The modeling for the pH and ionic strength effects on phosphate adsorption in soil generally falls within the error bars of parallel experiments (Figure 5). However, a significant deviation is observed for S1 and S2 under alkaline conditions.
The modeling results indicated that the increase in pH led to an increased saturation index of hydroxyapatite during the batch experiment, reaching saturation under alkaline conditions and resulting in the precipitation of hydroxyapatite (Figure S5), primarily contributed by the soluble Ca2+ in the soil. Nevertheless, even when considering the presence of soluble cations, the simulation results were unable to fully explain the retention of phosphate in S1 and S2 (Figure 5). Therefore, the enhanced adsorption of phosphate under alkaline conditions is not attributed to the soluble Ca2+.
To identify the specific minerals responsible for enhanced phosphate retention under alkaline conditions, the correlation coefficient between the maximum simulation error and the soil mineral content was calculated (Figure 7). The results revealed that all three soil samples exhibited the highest simulation errors under alkaline conditions, and these error values were strongly positively correlated with the content of biotite and vermiculite. This result suggests that the enhanced retention of phosphate in the soil under alkaline conditions is directly related to biotite or vermiculite.
To assess whether phosphate adsorption was enhanced by the leaching of iron elements from biotite, the FeEx content in the soil was determined both before and after phosphate adsorption under acidic and alkaline conditions. The results showed that the FeEx content of the soil did not significantly change under any conditions (Figure S6), indicating that alkaline conditions did not induce the leaching of iron elements from biotite or the formation of additional iron oxides.
Among the reported layered silicate minerals, montmorillonite exhibits a similar characteristic of enhanced phosphate adsorption under alkaline conditions as observed in S1 and S2, attributed to the leaching of Ca2+ from interlayer spaces, leading to the precipitation of phosphates [51]. Similarly, vermiculite also contains exchangeable cations in its interlayers, with hydrated Na+ and Ca2+ being common interlayer cations [52]. Biotite typically contains non-hydrated K+ within its interlayer spaces and has poorer exchangeability. However, in naturally weathered biotite, K+ at the edge fracture sites is often replaced by hydrated Na+ and Ca2+ [53]. Therefore, under the influence of high concentrations of Na+ in solution, both vermiculite and biotite have the potential to release Ca2+, which may then form precipitates with phosphate under alkaline conditions. The contents of vermiculite in S1, S2, and S3 are 1.17%, 0.49%, and 0.00%, respectively, consistent with the experimental observations of a larger adsorption anomaly in S1 compared to S2, and no significant adsorption anomaly in S3.
Thus, under batch experiment conditions, Ca2+ within the interlayer spaces of vermiculite is exchanged with Na+ from the liquid phase into the solution and reacts with the dissolved phosphate to form hydroxyapatite precipitate. The reaction can be represented by the following equation:
Ca2+(s) + 2 Na+(aq) = 2 Na+(s) + Ca2+(aq)
10 Ca2+(aq) + 6 PO43−(aq) + 2 OH−(aq) → Ca10(PO4)6(OH)2(s)
Specifically, cation exchange reactions are influenced by the substrate and are facilitated by an increase in Na+ concentration in the solution. Therefore, under otherwise identical conditions, the concentration of exchanged Ca2+ increases with the increasing concentration of Na+ in the liquid phase. The increase in Ca2+ concentration promotes the precipitation of hydroxyapatite, thereby enhancing the retention of phosphate. This is consistent with the ionic strength effect observed for S1 and S2 (Figure 5).
Unfortunately, the XRD patterns of the soil samples before and after phosphate treatment were largely unchanged, and no characteristic peaks of hydroxyapatite were observed in the soil samples treated with phosphate under either acidic or alkaline conditions (Figure S7). We note that even if the maximum adsorption capacity of 0.8 mg/g was entirely contributed by precipitation, the content of hydroxyapatite in the soil would be only 0.43% (Figure 5). Therefore, the low content of hydroxyapatite in the soil, possibly below the detection limit, may be the primary reason why it is not discernible by XRD.
Distinct P-O bands were observed in the FTIR spectra of both the soil before and after phosphate adsorption (Figure 8). The absorption peak at 1030 cm−1 is a typical indication of phosphate adsorption on ferrihydrite, while the peak at 1110 cm−1 is attributed to phosphate adsorption on goethite. The peaks at 1008 cm−1 and 1090 cm−1 are contributed by phosphate adsorption on both ferrihydrite and goethite [54]. These findings suggest that the predominant forms of free iron in the soil are ferrihydrite and goethite, both of which contribute to phosphate adsorption.
For S1 and S2, clear P-O bands were observed in both the original soil and the samples adsorbed with phosphate under acidic conditions, indicating the presence of phosphate adsorbed on iron oxides in both cases. In contrast, under alkaline conditions, the P-O bands in S1 and S2 significantly weakened, suggesting that phosphate was not present on the surface of iron oxides through adsorption, and that the adsorbed phosphate in the original soil was desorbed.
For S3, the content of adsorbed phosphate in the original soil was lower. The enhancements in the P-O bands were observed under both acidic and alkaline conditions, indicating that phosphates were consistently immobilized in the soil through adsorption. This observation is consistent with the previously noted phenomena (Figure 5).
Overall, the retention of phosphate in soil samples involves two mechanisms, including adsorption and precipitation. Under acidic conditions, phosphate is primarily adsorbed by free iron and aluminum oxides via forming inner-sphere surface complexes. In contrast, under alkaline conditions, the formation of hydroxyapatite induced by Ca2+ from soil soluble salts and layered silicate minerals dominates the immobilization of phosphate.

4.3. Retention Distribution of Phosphate in Soil

The retention distribution of phosphate in soil under different pH and ionic strength conditions was calculated based on experimental and modeling results. The amount of phosphate adsorbed by iron and aluminum oxides, as well as precipitated with soil soluble salts, was determined by the modeling result, while the amount of phosphate fixed by the release of Ca2+ from layered silicate minerals was determined by the positive error between experimental and modeling values. As shown in Figure 9, under acidic and neutral conditions, phosphate adsorption is predominantly dominated by iron oxides, contributing 80–90% of the soil phosphate retention. Only about 10% of the fixed phosphate is adsorbed by aluminum oxides, which can be attributed to the higher content of free iron in the soil, along with a greater specific surface area and surface site density (Table S1 and Table 4). Furthermore, the contribution from layered silicate minerals observed under acidic conditions is due to the uncertainty in batch experiments, as the presence of H+ is not conducive to the precipitation of hydroxyapatite.
When the solution shifts to alkaline conditions, phosphate dissociates into HPO42−, which exhibits a lower adsorption affinity, resulting in a reduction in the amount of phosphate adsorbed on iron and aluminum oxides. Upon reaching a critical pH, phosphate is predominantly removed from the liquid phase through the precipitation of hydroxyapatite. The contribution of soil soluble salts to phosphate retention also increases under alkaline conditions but remains below 10%, as the content of soluble Ca2+ in all three samples is relatively low. Notably, although the enhanced retention of phosphate in S3 under alkaline conditions is not significant, from a proportional perspective, S3 also demonstrates phosphate retention due to layered silicate minerals, with a critical pH of approximately 11.0. This phenomenon suggests that, in addition to vermiculite, biotite also contributes to the fixation of phosphate to some extent.
It is noteworthy that the critical pH for the formation of hydroxyapatite is not fixed and is influenced by both soil mineralogy and solution conditions. Specifically, an increase in the content of layer silicate minerals containing exchangeable Ca2+ in the soil will lead to a decreased critical pH, as this favors the enrichment of Ca2+ in the solution. As indicated by the reaction equation for the formation of hydroxyapatite, an increase in Ca2+ concentration will result in a lower equilibrium concentration of OH−, corresponding to a reduced critical pH. Similarly, solution conditions, particularly an increase in Na+ concentration, can enhance the exchange of Ca2+ into the solution, further lowering the critical pH. These inferences are consistent with the phenomena illustrated in Figure 9. Moreover, although there is no direct evidence, it is speculated that the type of cations in the solution, which possess higher valence states, smaller radii and thus stronger exchange capacities, may potentially influence the critical pH [55].

4.4. Environmental Implication

This study highlights the mineralogical control of phosphate adsorption in soil. Previous studies have demonstrated that phosphate is immobilized through adsorption of iron/aluminum oxides under acidic and neutral conditions, while under alkaline conditions, it mainly reacts with free Ca2+ or is stabilized on carbonate minerals by surface precipitation. In this study, we further reveal the significant role of non-free Ca2+. Specifically, under alkaline conditions, Ca2+ within the interlayer spaces of layered silicate minerals can be exchanged into the solution and immobilize phosphate by forming hydroxyapatite. As this study focused on only three soil samples, the impact of other minerals, such as montmorillonite, chlorite, kaolinite, and illite, on phosphate retention remains unexplored. Nevertheless, the significant role of vermiculite in phosphate retention has been confirmed, allowing the conclusions of this study to be extended to other soils and environments rich in vermiculite.
Since soil acts as a critical barrier to the entry of phosphate into aquifers, selecting sites that promote phosphate retention can effectively mitigate the environmental risks related industries pose to groundwater and downstream surface water. Generally, wastewater from the phosphate chemical industry is acidic, suggesting that soils with higher contents of iron/aluminum oxides exhibit superior phosphate retardation capabilities. For enterprises discharging alkaline phosphate-containing wastewater, such as those involved in alkaline nickel electroplating, it is advisable to select sites with higher contents of carbonate minerals, soluble Ca2+, and layered silicate minerals (especially those with high exchangeable Ca2+ content, such as montmorillonite and vermiculite). It is essential to recognize that alkaline soils can alkalize wastewater, negatively impacting phosphate adsorption on iron/aluminum oxides. Therefore, when establishing a phosphate-related enterprise at an alkaline site, it may be considered to blend a certain proportion of calcite, vermiculite, and montmorillonite into the soil to enhance its phosphate retention capacity.
Vermiculite is widely used in agriculture as a soil amendment to supply crops with essential trace elements [56]. Based on the findings of this study, we propose that vermiculite incorporation in soils also plays a significant role in mitigating agricultural non-point source pollution. When excessive fertilization occurs, phosphate reacts with exchangeable Ca2+ released from vermiculite to form precipitates, effectively retaining phosphate in the topsoil. This mechanism not only prevents groundwater phosphate pollution but also enhances soil fertility preservation.
As a low-cost natural mineral, vermiculite is priced at approximately 600 RMB (84 USD) per ton in China. Achieving a 10% vermiculite content in the top 20 cm soil layer would require an estimated amendment cost of 1300 RMB (186.7 USD) per acre (0.067 hectares), a cost that remains highly feasible for industrial applications. Notably, vermiculite can be sourced from phlogopite mining waste, further reducing implementation costs [57]. In agricultural applications, vermiculite effectively retains nutrients, thereby reducing fertilizer expenditures [56]. Collectively, these factors demonstrate the cost-effectiveness of vermiculite in both industrial and agricultural contexts.

5. Conclusions

The macroscopic adsorption behavior of phosphate in soil was investigated through batch adsorption experiments. Correlation between soil adsorption behavior and soil mineralogy enabled the identification of target minerals and potential retention mechanisms for phosphate in soil, which were subsequently verified using FTIR. A geochemical model was developed via the GAA method to quantitatively investigate the retention and distribution of phosphate within the soil samples. The key findings are as follows:
  • Phosphate adsorption in soil follows the intra-particle diffusion model, reaching equilibrium approximately 750 min after initiation. The adsorption capacity ranges from 0.193 to 0.217 mg/g. Phosphate adsorption is characterized by spontaneous endothermic reactions and partial desorption. An increase in pH typically inhibits phosphate adsorption, though it is enhanced under alkaline conditions in soils containing vermiculite. Ionic strength generally does not affect phosphate adsorption, but enhances adsorption in vermiculite-containing samples under alkaline conditions.
  • Phosphate is predominantly adsorbed on iron oxides in the form of inner-sphere surface complexes (approximately 80–90%) under acidic and neutral conditions. Aluminum oxides, due to their low content, smaller specific surface area, and lower surface site density, contribute only about 10% to the adsorbed phosphate in soil.
  • Under alkaline conditions, the affinity of iron/aluminum oxides for phosphate decreases, and layered silicate minerals become the primary contributors to phosphate retention. Specifically, the release of exchangeable Ca2+ from the interlayer spaces of vermiculite and biotite induces the precipitation of hydroxyapatite.
  • The critical pH for the transition of adsorption mechanisms decreases with the increasing content of vermiculite in soil and the concentration of cations in solution, and may also be potentially influenced by the type of cations present.
  • Given that soil serves as a primary barrier to phosphate entering into groundwater, it is recommended to select sites that are mineralogically favorable and align with the pH characteristics of the wastewater for phosphate-related factories, thereby minimizing the environmental risks to groundwater and downstream surface water. Incorporation of a specified quantity of vermiculite into soils effectively mitigates groundwater phosphate contamination and demonstrates significant cost-effectiveness in both industrial and agricultural applications.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/w17070998/s1, Figure S1: The XRD patterns of the soil samples. Figure S2: Comparison of experimental values and simulation results for the equilibrium constant of surface complexation reactions of phosphate on (a) free iron oxides [31] and (b) free aluminum oxides [32]. Figure S3: The change of the output adsorption capacity of the geochemical model when the input parameter changes by 10%. Figure S4: The grain size classification of the soil samples. Figure S5: The saturation index of hydroxyapatite in the batch experimental system. Figure S6: The FeEx contents of soil samples after different treatment. Figure S7: The XRD patterns of soil samples before and after phosphate adsorption. Table S1: Chemical and mineralogy compositions of the soil samples. Table S2: Content of soluble cations in soil samples. Table S3: Raw data for adsorption kinetic experiments. Table S4: Raw data for isothermal adsorption experiments at 25 °C. Table S5: Raw data for isothermal adsorption experiments at 45 °C. Table S6: Raw data for isothermal adsorption experiments at 60 °C. Table S7: Raw data for desorption experiments. Table S8: Raw data for the experiments of the impact of adsorbent dose. Table S9: Raw data for adsorption edges at 0.01 M NaCl electrolyte background. Table S10: Raw data for adsorption edges at 0.1 M NaCl electrolyte background. References [31,32] are cited in the Supplementary Materials.

Author Contributions

Conceptualization, Z.T. and Z.C.; Data curation: Z.T., Z.C. and N.X.; Formal analysis, Z.T. and F.J.; Funding acquisition: Z.T. and Z.C.; Investigation, Z.T., M.Z., S.F., L.W. and Q.F.; Methodology: Z.T., F.J., M.Z. and S.F.; Project administration: Z.T., Z.C. and N.X.; Resources: F.J., Y.W., Z.C. and N.X.; Software: Z.T., F.J. and Z.C.; Supervision: Z.C. and N.X.; Validation: Z.T. and Z.C.; Visualization: Z.T. and Z.C.; Writing—original draft: Z.T.; Writing—review & editing: Z.C. All authors have read and agreed to the published version of the manuscript.

Funding

This work was jointly supported by the Hubei Provincial Natural Science Foundation of China (No. 2023AFD222), the National Natural Science Foundation of China (42207237), the Xiaogan City Natural Science Research Program (No. XGKJ2024020004), the Natural Science Foundation of Jiangxi Province, China (No. 20242BAB20153), and the China Geological Survey Project (DD20243178).

Data Availability Statement

The original contributions presented in this study are included in the article/Supplementary Materials. Further inquiries can be directed to the corresponding authors.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Wang, Y.; Yuan, S.; Shi, J.; Ma, T.; Xie, X.; Deng, Y.; Du, Y.; Gan, Y.; Guo, Z.; Dong, Y.; et al. Groundwater Quantity and Health: Making the Invisible Visible. Environ. Sci. Technol. 2023, 57, 5125–5136. [Google Scholar] [PubMed]
  2. Warrack, J.; Kang, M.; von Sperber, C. Groundwater phosphorus concentrations: Global trends and links with agricultural and oil and gas activities. Environ. Res. Lett. 2022, 17, 014014. [Google Scholar]
  3. Huang, G.; Liu, C.; Zhang, Y.; Chen, Z. Groundwater is important for the geochemical cycling of phosphorus in rapidly urbanized areas: A case study in the Pearl River Delta. Environ. Pollut. 2020, 260, 114079. [Google Scholar] [PubMed]
  4. Chen, Y.; Ma, T.; Chen, L.; Liu, W.; Zhang, M.; Shang, R. Multimedia Nitrogen and Phosphorus Migration and Source Control Using Multivariate Analysis and XGBoost: The Case Study in a Typical Agricultural Basin, Danjiangkou Reservoir. Water 2024, 16, 1936. [Google Scholar] [CrossRef]
  5. Meinikmann, K.; Hupfer, M.; Lewandowski, J. Phosphorus in groundwater discharge—A potential source for lake eutrophication. J. Hydrol. 2015, 524, 214–226. [Google Scholar]
  6. Schoumans, O.F.; Chardon, W.J. Phosphate saturation degree and accumulation of phosphate in various soil types in The Netherlands. Geoderma 2015, 237–238, 325–335. [Google Scholar]
  7. Boschetti, T.; Hamed, Y.; Hadji, R.; Barbieri, M.; Gentilucci, M.; Rossi, M.; Khalil, R.; Khan, S.D.; Asghar, B.; Al-Omran, A.; et al. Using principal component analysis to distinguish sources of radioactivity and nitrates contamination in Southern Tunisian groundwater samples. J. Geochem. Explor. 2025, 271, 107670. [Google Scholar]
  8. Millán-Becerro, R.; Pérez-López, R.; Cánovas, C.R.; Macías, F.; León, R. Phosphogypsum weathering and implications for pollutant discharge into an estuary. J. Hydrol. 2023, 617, 128943. [Google Scholar]
  9. Melki, S.; El Asmi, A.M.; Gueddari, M. Inferred Industrial and Agricultural Activities Impact on Groundwater Quality of Skhira Coastal Phreatic Aquifer in Southeast of Tunisia (Mediterranean Region). Geofluids 2019, 1, 9465498. [Google Scholar]
  10. Melki, S.; Asmi, A.M.E.; Sy, M.; Gueddari, M.J.E.E.S. A geochemical assessment and modeling of industrial groundwater contamination by orthophosphate and fluoride in the Gabes-North aquifer, Tunisia. Environ. Earth Sci. 2020, 79, 1–20. [Google Scholar]
  11. Ahmed, A.A.; Morshedizad, M.; Kühn, O.; Leinweber, P. Deciphering competitive interactions: Phosphate and organic matter binding on goethite through experimental and theoretical insights. Sci. Total Environ. 2024, 940, 173510. [Google Scholar] [PubMed]
  12. Abdelwaheb, M.; Jebali, K.; Dhaouadi, H.; Dridi-Dhaouadi, S. Adsorption of nitrate, phosphate, nickel and lead on soils: Risk of groundwater contamination. Ecotoxicol. Environ. Saf. 2019, 179, 182–187. [Google Scholar] [PubMed]
  13. Haile, A.M.; Abera, T.; Molla, B.D.; Tamiru, A.A.; Oddvar, R. Surface-complexation modelling for describing adsorption of phosphate on hydrous ferric oxide surface. Water SA 2015, 41, 157–166. [Google Scholar]
  14. Sarkar, B.; Mukhopadhyay, R.; Ramanayaka, S.; Bolan, N.; Ok, Y.S. The role of soils in the disposition, sequestration and decontamination of environmental contaminants. Philos. Trans. R. Soc. B-Biol. Sci. 2021, 376, 20200177. [Google Scholar]
  15. Ahmed, A.A.; Leinweber, P.; Kühn, O. Advances in understanding the phosphate binding to soil constituents: A Computational Chemistry perspective. Sci. Total Environ. 2023, 887, 163692. [Google Scholar] [CrossRef]
  16. Rout, P.R.; Bhunia, P.; Dash, R.R. A mechanistic approach to evaluate the effectiveness of red soil as a natural adsorbent for phosphate removal from wastewater. Desalin. Water Treat. 2015, 54, 358–373. [Google Scholar]
  17. Mª García-López, A.; Delgado, A.; Plassard, C. Kinetics of phytate adsorption and response of phosphorus forms initially present in alkaline soils. Geoderma 2024, 443, 116800. [Google Scholar]
  18. Huang, J.; Huang, X.; Jiang, D. Phosphorus can effectively reduce selenium adsorption in selenium-rich lateritic red soil. Sci. Total Environ. 2024, 906, 167356. [Google Scholar]
  19. Amarh, F.; Voegborlo, R.B.; Essuman, E.K.; Agorku, E.S.; Tettey, C.O.; Kortei, N.K. Effects of soil depth and characteristics on phosphorus adsorption isotherms of different land utilization types: Phosphorus adsorption isotherms of soil. Soil Tillage Res. 2021, 213, 105139. [Google Scholar]
  20. Wang, Y.; Jiang, J.; Xu, R.K.; Tiwari, D. Phosphate adsorption at variable charge soil/water interfaces as influenced by ionic strength. Aust. J. Soil Res. 2009, 47, 529–536. [Google Scholar] [CrossRef]
  21. Geng, Y.; Pan, S.; Zhang, L.; Qiu, J.; He, K.; Gao, H.; Li, Z.; Tian, D. Phosphorus biogeochemistry regulated by carbonates in soil. Environ. Res. 2022, 214, 113894. [Google Scholar] [CrossRef]
  22. Gustafsson, J.P.; Mwamila, L.B.; Kergoat, K. The pH dependence of phosphate sorption and desorption in Swedish agricultural soils. Geoderma 2012, 189–190, 304–311. [Google Scholar]
  23. Tuntrachanida, J.; Wisawapipat, W.; Aramrak, S.; Chittamart, N.; Klysubun, W.; Amonpattaratkit, P.; Duboc, O.; Wenzel, W.W. Combining spectroscopic and flux measurement techniques to determine solid-phase speciation and solubility of phosphorus in agricultural soils. Geoderma 2022, 410, 115677. [Google Scholar]
  24. Zhao, W.; Gu, C.; Zhu, M.; Yan, Y.; Liu, Z.; Feng, X.; Wang, X. Chemical speciation of phosphorus in farmland soils and soil aggregates around mining areas. Geoderma 2023, 433, 116465. [Google Scholar]
  25. Pechenyuk, S.I.; Semushina, Y.P.; Kuz’mich, L.F. Adsorption affinity of anions on metal oxyhydroxides. Russ. J. Phys. Chem. A 2013, 87, 490–496. [Google Scholar]
  26. Gérard, F. Clay minerals, iron/aluminum oxides, and their contribution to phosphate sorption in soils—A myth revisited. Geoderma 2016, 262, 213–226. [Google Scholar]
  27. Jin, J.; Xiong, J.; Liang, Y.; Wang, M.; Huang, C.; Koopal, L.; Tan, W. Generic phosphate affinity constants of the CD-MUSIC-eSGC model to predict phosphate adsorption and dominant speciation on iron (hydr)oxides. Water Res. 2024, 264, 122194. [Google Scholar]
  28. Nam, G.; Choi, Y.H.; Lee, N.; Ahn, J.W. Effect by Alkaline Flocculation of Algae and Phosphorous from Water Using a Calcined Waste Oyster Shell. Water 2017, 9, 661. [Google Scholar] [CrossRef]
  29. Fan, B.; Ding, J.; Fenton, O.; Daly, K.; Chen, Q. Understanding phosphate sorption characteristics of mineral amendments in relation to stabilising high legacy P calcareous soil. Environ. Pollut. 2020, 261, 114175. [Google Scholar]
  30. Sø, H.U.; Postma, D.; Jakobsen, R.; Larsen, F. Sorption of phosphate onto calcite; results from batch experiments and surface complexation modeling. Geochim. Cosmochim. Acta 2011, 75, 2911–2923. [Google Scholar]
  31. Antelo, J.; Fiol, S.; Pérez, C.; Mariño, S.; Arce, F.; Gondar, D.; López, R. Analysis of phosphate adsorption onto ferrihydrite using the CD-MUSIC model. J. Colloid Interface Sci. 2010, 347, 112–119. [Google Scholar] [CrossRef] [PubMed]
  32. Laiti, E.; Persson, P.; Öhman, L.O. Balance between Surface Complexation and Surface Phase Transformation at the Alumina/Water Interface. Langmuir 1998, 14, 825–831. [Google Scholar] [CrossRef]
  33. Bullen, J.C.; Landa-Cansigno, O.; Weiss, D.J. Teaching Adsorption Chemistry by Constructing Surface Complexation Models (SCM) in PHREEQC. J. Chem. Educ. 2024, 101, 1914–1924. [Google Scholar] [CrossRef]
  34. Huang, L.; Fang, H.; Reible, D. Mathematical model for interactions and transport of phosphorus and sediment in the Three Gorges Reservoir. Water Res. 2015, 85, 393–403. [Google Scholar] [CrossRef]
  35. Davis, J.A.; Coston, J.A.; Kent, D.B.; Fuller, C.C. Application of the surface complexation concept to complex mineral assemblages. Environ. Sci. Technol. 1998, 32, 2820–2828. [Google Scholar] [CrossRef]
  36. HJ 889-2017; Soil Quantity—Determination of Cation Exchange Capacity (CEC)—Hexamminecobalt Trichloride Solution-Spectrophotometric Method. Ministry of Ecology and Environment of China: Beijing, China, 2017.
  37. Chi, Z.; Xie, X.; Pi, K.; Wang, Y. Mineralogical controls on arsenite adsorption onto soils: Batch experiments and model-based quantification. Sci. Total Environ. 2021, 767, 144920. [Google Scholar] [CrossRef]
  38. Pérez, C.; Antelo, J.; Fiol, S.; Arce, F. Modeling oxyanion adsorption on ferralic soil, part 1: Parameter validation with phosphate ion. Environ. Toxicol. Chem. 2014, 33, 2208–2216. [Google Scholar] [CrossRef]
  39. Wang, J.; Guo, X. Adsorption kinetic models: Physical meanings, applications, and solving methods. J. Hazard. Mater. 2020, 390, 122156. [Google Scholar] [CrossRef]
  40. Qing, Z.; Wang, L.; Liu, X.; Song, Z.; Qian, F.; Song, Y. Simply synthesized sodium alginate/zirconium hydrogel as adsorbent for phosphate adsorption from aqueous solution: Performance and mechanisms. Chemosphere 2022, 291, 133103. [Google Scholar] [CrossRef]
  41. Kalam, S.; Abu-Khamsin, S.A.; Kamal, M.S.; Patil, S. Surfactant Adsorption Isotherms: A Review. ACS Omega 2021, 6, 32342–32348. [Google Scholar] [CrossRef]
  42. Samimi, M.; Zakeri, M.; Alobaid, F.; Aghel, B. A Brief Review of Recent Results in Arsenic Adsorption Process from Aquatic Environments by Metal-Organic Frameworks: Classification Based on Kinetics, Isotherms and Thermodynamics Behaviors. Nanomaterials 2023, 13, 60. [Google Scholar] [CrossRef] [PubMed]
  43. Qiu, T.; Zeng, Y.; Ye, C.; Tian, H. Adsorption Thermodynamics and Kinetics of p-Xylene on Activated Carbon. J. Chem. Eng. Data 2012, 57, 1551–1556. [Google Scholar] [CrossRef]
  44. Oh, Y.M.; Hesterberg, D.L.; Nelson, P.V.; Niedziela, C.E., Jr. Desorption Characteristics of Three Mineral Oxides and a Non-crystalline Aluminosilicate for Supplying Phosphate in Soilless Root Media. Commun. Soil Sci. Plant Anal. 2016, 47, 753–760. [Google Scholar] [CrossRef]
  45. Wang, F.; Xu, J.; Xu, Y.; Chen, H.; Liang, Y.; Xiong, J. Face-dependent phosphate speciation on goethite: CD-MUSIC modeling and ATR-FTIR/2D-COS study. Sci. Total Environ. 2024, 915, 169970. [Google Scholar] [CrossRef]
  46. Ioannou, A.; Dimirkou, A. Phosphate Adsorption on Hematite, Kaolinite, and Kaolinite–Hematite (k–h) Systems As Described by a Constant Capacitance Model. J. Colloid Interface Sci. 1997, 192, 119–128. [Google Scholar] [CrossRef]
  47. Chi, Z.; Xie, X.; Pi, K.; Wu, Y.; Wang, Y. Spectroscopic and modeling approaches of arsenic (III/V) adsorption onto Illite. J. Hazard. Mater. 2024, 477, 135284. [Google Scholar]
  48. Gustafsson, J.P.; Barbi, F.; Braun, S.; Klysubun, W.; Börjesson, G. Where does all the phosphorus go? Mass balance modelling of phosphorus in the Swedish long-term soil fertility experiments. Geoderma 2023, 440, 116727. [Google Scholar] [CrossRef]
  49. Liang, H.; Liu, J.; Wei, Y.; Guo, X. Evaluation of phosphorus removal from wastewater by soils in rural areas in China. J. Environ. Sci. 2010, 22, 15–22. [Google Scholar] [CrossRef]
  50. Tao, Y.; Du, Y.; Deng, Y.; Huang, Y.; Leng, Z.; Ma, T.; Wang, Y. Carbon and iron isotope approach elucidates the enrichment of geogenic phosphorus in alluvial-lacustrine sedimentary aquifers. J. Hydrol. 2022, 607, 127517. [Google Scholar] [CrossRef]
  51. Edzwald, J.K.; Toensing, D.C.; Leung, M.C.Y. Phosphate adsorption reactions with clay minerals. Environ. Sci. Technol. 1976, 10, 485–490. [Google Scholar] [CrossRef]
  52. do Nascimento, F.H.; Masini, J.C. Vermiculite Modified with Fe3+ Polyhydroxy Cations Is a Low-Cost and Highly Available Adsorbent for the Removal of Phosphate Ions. Minerals 2022, 12, 1033. [Google Scholar] [CrossRef]
  53. Kyllönen, J.; Hakanen, M.; Lindberg, A.; Harjula, R.; Vehkamäki, M.; Lehto, J. Modeling of cesium sorption on biotite using cation exchange selectivity coefficients. Radiochim. Acta 2014, 102, 919–929. [Google Scholar]
  54. Krumina, L.; Kenney, J.P.L.; Loring, J.S.; Persson, P. Desorption mechanisms of phosphate from ferrihydrite and goethite surfaces. Chem. Geol. 2016, 427, 54–64. [Google Scholar]
  55. Huang, Y.; Zhang, Z. Ion exchange selectivity (Mg2+, Ca2+ and K+) in hydrated Na-montmorillonite: Insights from molecular dynamic simulations. Mol. Simul. 2023, 49, 223–232. [Google Scholar]
  56. Zhang, H.; Yao, Y. Vermiculite addition to soil decreases N water pollution by over 30%. Environ. Chem. Lett. 2017, 15, 507–513. [Google Scholar]
  57. Kremenetskaya, I.; Tereshchenko, S.; Alekseeva, S.; Mosendz, I.; Slukovskaya, M.; Ivanova, L.; Mikhailova, I. Vermiculite-lizardite ameliorants from mining waste. IOP Conf. Ser. Earth Environ. Sci. 2019, 368, 012027. [Google Scholar]
Water 17 00998 g001
Figure 1. Location of the Huangmailing phosphate mine and the soil samples.
Figure 1. Location of the Huangmailing phosphate mine and the soil samples.
Water 17 00998 g001
Water 17 00998 g002
Figure 2. Results of (A–C) Langmuir, Sips, and (D–F) Freundlich fitting for the isotherms of phosphate adsorption onto soil samples. The red lines in (A–C) represent the modeling results of geochemical model.
Figure 2. Results of (A–C) Langmuir, Sips, and (D–F) Freundlich fitting for the isotherms of phosphate adsorption onto soil samples. The red lines in (A–C) represent the modeling results of geochemical model.
Water 17 00998 g002
Water 17 00998 g003
Figure 3. Comparison of adsorption and desorption curves of phosphate in soil samples.
Figure 3. Comparison of adsorption and desorption curves of phosphate in soil samples.
Water 17 00998 g003
Water 17 00998 g004
Figure 4. Impact of soil dose on the adsorption of phosphate onto soil samples.
Figure 4. Impact of soil dose on the adsorption of phosphate onto soil samples.
Water 17 00998 g004
Water 17 00998 g005
Figure 5. Impact of pH and ion strength on the adsorption of phosphate onto soil samples.
Figure 5. Impact of pH and ion strength on the adsorption of phosphate onto soil samples.
Water 17 00998 g005
Water 17 00998 g006
Figure 6. Correlation coefficient between soil mineral content and phosphate adsorption capacity.
Figure 6. Correlation coefficient between soil mineral content and phosphate adsorption capacity.
Water 17 00998 g006
Water 17 00998 g007
Figure 7. Correlation coefficient between soil mineral content and maximum simulation error.
Figure 7. Correlation coefficient between soil mineral content and maximum simulation error.
Water 17 00998 g007
Water 17 00998 g008
Figure 8. Comparison of FTIR spectra of soil samples before and after phosphate adsorption under acidic and alkaline conditions.
Figure 8. Comparison of FTIR spectra of soil samples before and after phosphate adsorption under acidic and alkaline conditions.
Water 17 00998 g008
Water 17 00998 g009
Figure 9. The effects of pH and ionic strength on the retention distribution of phosphate in soil samples.
Figure 9. The effects of pH and ionic strength on the retention distribution of phosphate in soil samples.
Water 17 00998 g009
Table 1. The specific experimental setup of adsorption batch experiments.
Table 1. The specific experimental setup of adsorption batch experiments.
Temperature
(°C)		Time
(h)		Solid to Liquid Ratio
(g/mL)		Electrolyte Backgrounds
(M NaCl)		Initial pHInitial Phosphate Concentration
(mg/L)
Adsorption kinetic250–3340:5000.017.050
Adsorption isotherm25242:250.017.05/15/30/45/70/100
Adsorption Thermodynamics45242:250.017.05/15/30/45/70/100
60242:250.017.05/15/30/45/70/100
Desorption25122:25/31.25/41.67/62.50.017125
Adsorbent dose25240.5/1/1.5/2/3:250.017.0100
Adsorption edge25242:250.012.5–12100
25242:250.12.5–12100
Table 2. Fitted kinetic parameters for phosphate adsorption on soil samples.
Table 2. Fitted kinetic parameters for phosphate adsorption on soil samples.
Kinetic ModelParametersSoil Sample
S1S2S3
Pseudo-first-orderK1 (1/min)0.03260.01020.0554
Qmax (mg/g)0.06190.13360.0740
R20.8100.7780.666
Pseudo-second-orderK2 [g/(mg·min)]0.67150.09860.9210
Qmax (mg/g)0.06590.14410.0769
R20.9190.8920.820
Intra-particle diffusionKd1 [mg/(g·min1/2)]0.02670.02500.0108
R20.9490.9990.999
Kd2 [mg/(g·min1/2)]0.00780.02250.0117
R20.9981.0001.000
Kd3 [mg/(g·min1/2)]0.00590.01450.0031
R20.0070.9850.744
Table 3. Thermodynamic parameters of phosphate adsorption onto soil samples.
Table 3. Thermodynamic parameters of phosphate adsorption onto soil samples.
T (°C)ΔG0 (KJ/mol)ΔS0 (KJ/mol/K)ΔH0 (KJ/mol)
S125−10.3700.15335.18
45−13.5770.153
60−15.6960.153
S225−8.420.14914.61
45−10.500.147
60−11.040.142
S325−10.490.1563.93
45−13.120.155
60−11.940.144
Table 4. Adsorption of phosphate in iron oxides and aluminum oxides.
Table 4. Adsorption of phosphate in iron oxides and aluminum oxides.
MineralsSite DensitySurface AreaSurface Complexation ReactionsLog K
Fe minerals3.45 site/nm2300 m2/gFe_sOH + H+ = Fe_sOH2+2.80
Fe_sOH = Fe_sO− + H+−10.40
2Fe_sOH + 3H+ + PO43− = Fe_s2HPO4 + 2H2O28.50
2Fe_sOH + 2H+ + PO43− = Fe_s2PO4− + 2H2O23.20
Fe_sOH + H+ + PO43− = Fe_sPO42− + H2O14.80
Al minerals2.00 site/nm2100 m2/gAl_sOH + H+ = Al_sOH2+3.20
Al_sOH = Al_sO− + H+−9.05
Al_sOH + 2H+ + PO43− = Al_sHPO4− + H2O24.90
Al_sOH + H+ + PO43− = Al_sPO42− + H2O16.80
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Tang, Z.; Chi, Z.; Jiang, F.; Zhao, M.; Fu, S.; Wei, L.; Feng, Q.; Wu, Y.; Xu, N. Mechanisms and Implications of Phosphate Retention in Soils: Insights from Batch Adsorption Experiments and Geochemical Modeling. Water 2025, 17, 998. https://doi.org/10.3390/w17070998

AMA Style

Tang Z, Chi Z, Jiang F, Zhao M, Fu S, Wei L, Feng Q, Wu Y, Xu N. Mechanisms and Implications of Phosphate Retention in Soils: Insights from Batch Adsorption Experiments and Geochemical Modeling. Water. 2025; 17(7):998. https://doi.org/10.3390/w17070998

Chicago/Turabian Style

Tang, Zhi, Zeyong Chi, Fengcheng Jiang, Mingzhe Zhao, Shengbo Fu, Lingqiao Wei, Qingsheng Feng, Yongming Wu, and Nuchao Xu. 2025. "Mechanisms and Implications of Phosphate Retention in Soils: Insights from Batch Adsorption Experiments and Geochemical Modeling" Water 17, no. 7: 998. https://doi.org/10.3390/w17070998

APA Style

Tang, Z., Chi, Z., Jiang, F., Zhao, M., Fu, S., Wei, L., Feng, Q., Wu, Y., & Xu, N. (2025). Mechanisms and Implications of Phosphate Retention in Soils: Insights from Batch Adsorption Experiments and Geochemical Modeling. Water, 17(7), 998. https://doi.org/10.3390/w17070998

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop